The Legacy of Ancient Philosophies
What Classical Philosophies Can Teach Modern Artificial Intelligence
Source: https://pxhere.com
Since the dawn of humanity, philosophers have tried to search for wisdom and examine the origin and validity of human ideas, and values in an attempt to understand the universe, make decisions, and solve problems. Today, some principles can be explained by tracing the development of philosophy over the centuries. In the 21st century in particular, where artificial intelligence is considered one of the primary enabling tools, around the world, researchers became more interested in examining the ethics of these intelligent systems, and despite these efforts, it is still in its early stages. This article explores how ancient philosophies can inspire today's AI rules.
The Moral Machine Project (1) outlines scenarios that could occur when driving self-driving cars. One scenario asks about a driver who must choose one of 2 routes. The first will lead to crashing into a concrete barrier that will kill the driver, and the second is to continue ahead and hit pedestrians crossing the street who happen to be criminals. Should the driver protect him/herself or the pedestrians who happen to be criminals?
Philosophers in the early ages did not agree on a single answer to controversial questions, such as those that investigate the ethics of killing evil figures during their rise to power, and when such a question was posed to the Dalai Lama, he said: Yes, it is moral to kill that person, but do not get angry. (2) In Buddhism, morality is a fundamental principle, and therefore killing a person in an attempt to reduce the suffering of the world will remain an ethical act. This reflects the Buddhist vision that doesn't revolve around the person but to the good that individuals can do for the world as a whole.
Let's change the scenario a little and imagine a child crossing the street instead of criminals. It will remain difficult to make such a decision, and unlike Buddhism, which doesn't revolve around you as an individual, Confucianism teaches us self-recognition as part of our relationships with others, so regardless of which decision the driver chooses, he/she will still feel worried about the child at risk and that comes from his/her realization that we are all deeply connected.
You may have heard of ProPublica's case (3), which highlights an artificial intelligence system that imposed harsher sentences on black people compared to white people. If this case occurred in the golden age of Aristotelianism, it would be interpreted by one of its principles that believe in the role that external factors play - such as being born with certain skin color, ethnicity, and wealth - in the granting or denial of privileges that increase or decrease the potential of someone succeeding in life. Stoics also embraces the same idea. But it says in addition to that, that the level of someone's wealth does not affect how good they are, that is, wealth and evil can meet in the same person.
It seems clear from the previous examples that philosophers did not unite on one philosophy and interpretation, but rather some were contradictory to the other, and this comes from the fact that they came from different dimensions that reflect individual experiences and interests, and despite recent attempts to find a global model for the ethics of artificial intelligence, classical philosophies have taught us to this day that there are no universal answers to ethical questions and the ideal global model for AI ethics might be the one that allows for multiple models.
More:
1. https://www.moralmachine.net
2. https://www.amazon.com/How-Live-Good-Life-Philosophy/dp/0525566147
3. https://www.propublica.org/article/machine-bias-risk-assessments-in-criminal-sentencing